Quantile normalization

Quantile normalization is a mathematical averaging technique that allows comparison of different samples run on the same microarray chip. This is important because the probe intensity can vary from assay to assay due to technical differences that are clinically irrelevant. The cartoon illustrates probe intensities from four patients (1, 2, 3, and 4) run on the same type of chip. Most chips contain thousands to millions of individual probes. The Y axis shows the intensity of the probes, depicted as a full range (vertical blue lines), the interquartile range (black box), and the mean (horizontal green lines).
(A) Before normalization. The probe intensities vary from chip to chip.
(B) After normalization. The distribution of intensities and interquartile ranges are nearly identical, with small changes in the means. This allows proper analysis of relative gene expression from the four different patients.